does utilitarianism require sophistry?

Here's an interview with Peter Singer about his new book and about the Peter Singer position generally. A lot of it is a rehash of typical objections and responses, but there was a bit at the beginning that got me thinking:

"I don’t think I’ve diluted or changed anything of substance,” says Singer, who is unapologetic about the book’s populist style, which favours dealing with objections from students at Glennview High rather than professors at Harvard and Oxford. “I want it to be read by more than professors and students of philosophy. I do want to make a difference to people, I want to persuade as many people as possible to do something about world poverty. So I’ve got to reach a wide audience."

The usual norms governing philosophical writing---at least among professional academics like Singer---include "write what you believe to be correct" and "provide what you take to be good arguments." (As well as a bunch of others about brevity and relevance and what have you.) But for someone like Singer, whose impartial utilitarianism is unlimited in scope, moral norms might well override these norms of professional discourse, since writing and publishing philosophy are acts like any others. And so he might in fact be required to argue for something he doesn't believe is correct, employing arguments he doesn't take to be good, all for the sake of convincing others to do things which conduce to the aggregate good. I.e. utilitarianism might require sophistry.

In fact, I'm pretty sure it does, at least in some cases. Which makes me wonder why Peter Singer hasn't been engaging in it from the get-go. (Or maybe he has. But you'd think he'd do things differently if he were trying to change people's behavior...)